Infrastructure/Meetings/2006-03-02

Meeting of 2006-03-02

*** Time shown in CST

[14:05:49]   <mmcgrath>  Alllrighty, we ready to start?
[14:06:12]   <xDamox> yea
[14:06:13]   <mmcgrath> Who's here?
[14:06:14]   <mmcgrath> I am.
[14:06:19]   <ja8sun> ready
[14:06:20]   <xDamox> me
[14:06:47]   <mmcgrath> nman64 iwolf: we're starting without sopwith, you guys still here?
[14:06:47]   <kevinverma> I am
[14:07:23]   <mmcgrath> Item one on the agenda is the systems upgrades.
[14:07:37]   <mmcgrath> I've been working with Sopwith on this for Proxy1. We've hit some bumps.
[14:07:51]   <ja8sun> just got access to proxy3
[14:07:57]   <mmcgrath> Right now we're waiting on Stacy to get PXE setup I think so we can pxeboot that machine.
[14:08:10]   <nman64> I've been extremely busy, so I'm waiting and watching your efforts on proxy1 before I tackle app2.
[14:08:16]   <mmcgrath> Make sure you know whether the machine is using lilo or grub before you boot. I thought mine was using grub but it turned out to be lilo
[14:08:19]   <f13> mmcgrath: I'm sortof here.
[14:08:25]   <mmcgrath> f13: k
[14:08:38]   <iWolf> Back again.
[14:08:44]   <mmcgrath> welcome back.
[14:09:02]   <mmcgrath> Did everyone have a chance to look at my kickstart scripts?
[14:09:07]   <ja8sun> yes
[14:09:26]   <iWolf> Yep.
[14:09:30]   <xDamox> I need to have a look over it just uni have seem to through alot of assignments out this year :(
[14:09:32]   <mmcgrath> Personally I'm still apathetic as to whether or not to use a ks or whether to just use a VNC install.
[14:09:36]   <xDamox> but ill hav a look overit
[14:09:49]   <ja8sun> are we going to decide on what packages to install as a base
[14:10:13]   <iWolf> I think a set base might be good.  That way we know what we can count on being on every machine.
[14:10:15]   <mmcgrath> I'm actually leaning towards a VNC install because we don't really know how good a shape these machines are in and an interactive install might be more useful as far as troubleshooting goes if the install runs into any problems.
[14:10:20]   <iWolf> And then add based on the role of the server.
[14:10:31]   <iWolf> I also lean towards the VNC install.
[14:10:49]   <mmcgrath> but Its really up to you guys and your individual needs. I'm fine with a minimal install, get rid of all the packages we dont' need (as are listed in the ks script) and then adding packagese from there.
[14:10:59]   <xDamox> mmcgrath, yea VNC sounds good
[14:11:03]   <ja8sun> mmcgrath: VNC install sounds good.
[14:11:05]   <mmcgrath> err not really individual need but preference.
[14:11:24]   <kevinverma> rule no.1 on of SA always install what you need exclude rest
[14:11:27]   <mmcgrath> k, VNC it is. We can always revisit ks later but we're doing 10's of installs, not 100's.
[14:12:04]   <xDamox> mmcgrath, I am sure each of us could take a machine
[14:12:06]   <mmcgrath> Feel free to get started on your installs if you want. Mine's kind of on hold right now.
[14:12:14]   <mmcgrath> I think most people already have a machine from the last meeting.
[14:12:22]   <iWolf> mmcgrath: Are you documenting your steps so far?
[14:12:23]   <xDamox> I never got one :(
[14:12:48]   <iWolf> mmcgrath: So we don't work over the same hurdles?
[14:13:02]   <mmcgrath> xDamonx I think there's probably some to go around. Take a look at last weeks meeting and see which boxes haven't been claimed yet.
[14:13:12]   <xDamox> ok
[14:13:24]   <mmcgrath> I've been mindful of the steps so far but haven't actually documented them. I'll post to the list when we get a successful install.
[14:13:50]   <mmcgrath> A big problem I've been having is I can't get the cyclades serial interface to work correctly...
[14:14:01]   <ja8sun> mmcgrath:  that would be great!  Do we a place to store documentation?
[14:14:08]   <iWolf> Sounds good.  That will hopefully keep us from hitting the same roadblocks along the way and maybe make it easier in the future.
[14:14:25]   <mmcgrath> The wiki for the most part I guess.
[14:14:39]   <mmcgrath> though some of it might be considered 'security sensitive' passwords and things for example.
[14:14:47]   <mmcgrath> I'll talk to sopwith after we get my box installed.
[14:14:50]   <ja8sun> mmcgrath: wasn't sure if the wiki would be proper
[14:15:08]   <iWolf+ I would have thought the wiki a little too public.
[14:15:31]   <ja8sun^ That was my thought as well
[14:15:34]   <mmcgrath> We wouldn't store passwords out there. We also have documentation in the cvs that you guys could look at.
[14:15:55]   <ja8sun> mmcgrath: what module
[14:16:06]   * mmcgrath checking...
[14:16:07]   <iWolf+ that may be the best place to store build steps and such as well.
[14:16:35]   <mmcgrath> its under fedora-config
[14:16:51]   <ja8sun> mmcgrath: thanks, I'll take a look
[14:16:52]   <mmcgrath> there's a kickstart dir that has an "auto-ks" script.
[14:17:29]   <mmcgrath> that auto-ks script is a good starting point to see where to put the kernel, initrd.img and how to setup lilo or grub to boot to the installer (just remember to append vnc to the kernel params)
[14:17:42]   <mmcgrath> anywho, next subject: OTRS.
[14:18:02]   <mmcgrath> If anyone here has reviewer powers in extras please review OTRS for me, I'm still waiting on someone in Extras.
[14:18:15]   <mmcgrath> I'm going to try to put emphasis on it for Sunday's "review day" but we'll see.
[14:18:30]   <mmcgrath> The problem is no one is really familiar with it and people don't tend to like reviewing things they've never used before.
[14:18:37]   <mmcgrath> So thats still on hold.
[14:19:08]   <mmcgrath> The next item is seth's redhat boxes.
[14:19:29]   <mmcgrath> pining skvidal :-)
[14:19:57]   ***  skvidal has joined chat #fedora-admin.
[14:19:58]   <mmcgrath> he's comin
[14:20:04]   <skvidal> what's up?
[14:20:21]   <mmcgrath> skvidal: we're just going down the list here, I was wondering how the integration is going of your boxes and the other fedora boxes.
[14:20:34]   <skvidal> fpserv.linux.duke.edu is integrated
[14:20:42]   <mmcgrath> still waiting on extras64 though?
[14:20:49]   <skvidal> fpserv is supposed to replace fedora.linux.duke.edu - we were only waiting on the plan from the website folks
[14:21:05]   <skvidal> extras64 was put on hold pending FC5 release
[14:21:17]   <skvidal> I didn't think it was wise to mess with it while builds are ongoing
[14:21:24]   <skvidal> I thought sopwith and you agreed, too.
[14:21:27]   <skvidal> was I mistaken?
[14:21:38]   <mmcgrath> No I think thats good, I misread the notes.
[14:21:43]   <skvidal> okie doke
[14:21:59]   <skvidal> I would like to have fpserv.linux.duke.edu running sooner than later
[14:22:09]   <skvidal> if only b/c fedora.linux.duke.edu is a little aged
[14:22:20]   <mmcgrath> yeah, is fpserv just sitting there now not doing anything?
[14:22:22]   <skvidal> fpserv is up and happy - just waiting to know what the hell we have to install on it
[14:22:24]   <skvidal> yep
[14:22:28]   <skvidal> sitting there looking pretty
[14:22:30]   <skvidal> it's on
[14:22:32]   <skvidal> you can login
[14:22:38]   <skvidal> it occasionally makes annoying cron noises
[14:23:00]   <xDamox> heh
[14:23:01]   <mmcgrath> heh. what all needs to be done on it? Anything we can help with?
[14:23:12]   <skvidal> I don't have anything to do on it
[14:23:29]   <skvidal> I was told that the website folks were still making a decision about cms stuff
[14:23:38]   <skvidal> and that we would install whatever they finally settled on
[14:23:39]   <mmcgrath> seems we have a lot machines like that right now. Yeah, last I checked they still don't know.
[14:23:46]   <skvidal> ooooookie doke
[14:23:50]   <skvidal> so two options
[14:24:10]   <skvidal> 1. we just migrate things as they are from fedora.linux.duke.edu to fpserv.l.d.e
[14:24:18]   <skvidal> 2. we continue to wait for the website folks
[14:24:41]   <mmcgrath> I'd say 2. Its working fine as it is right now so there's no immediate need to migrate right?
[14:24:48]   <xDamox> I would suggest waiting for the web folks
[14:24:53]   <skvidal> yah - we can
[14:24:58]   <skvidal> but that has to have an upper limit
[14:25:03]   <skvidal> keep in mind
[14:25:03]   <xDamox> just incase they bitch about the migration
[14:25:12]   <skvidal> fedora.l.d.e is not new
[14:25:19]   <skvidal> it's a dual 1ghz with 1gb of ram
[14:25:22]   <mmcgrath> yeah. Maybe we should light a fire under their butts. Who's the main contact over there?
[14:25:28]   <mmcgrath> huzza.
[14:25:36]   <skvidal> I thought it was nman64 or maybe quaid
[14:25:43]   <mmcgrath> I thought nman64 too.
[14:25:55]   <mmcgrath> he was here earlier, maybe he'll be back. I'll ask him next time I see him.
[14:25:58]   * nman64 awakens.
[14:26:08]   <mmcgrath> there he is.
[14:26:34]   <mmcgrath> nman64: whats the word on the new cms system?
[14:26:51]   <nman64> The Websites team wants to stage Zope+Plone to start testing.
[14:27:21]   <mmcgrath> is it a proof of concept or do you want us to set up a place on the permanent server?
[14:27:58]   <nman64> We simply need a server that is ready for the job and packages built for the correct distribution release.  The goal is testing right now.  We can put it on a permanent location, but it would be easy to move later.
[14:28:36]   <mmcgrath> well its permanent location currently has nothing on it. So I'd assume if we put it there all we'd have to do before you guys went live is refresh any templates and wipe the database?
[14:28:54]   <nman64> I don't think we'd even have to do that much.
[14:29:22]   <mmcgrath> I'll add it to the schedule. Hopefully we'll get it done soon. skvidal: do you want to do this or should someone else?
[14:29:27]   <nman64> If all goes well, we'd just have to flip the proverbial switch.
[14:29:57]   <skvidal> mmcgrath: you can add it to the schedule
[14:30:04]   <xDamox> damn got to go Ill post the progress of Pootypedia on the mailing list
[14:30:09]   <xDamox> is that ok?
[14:30:13]   <mmcgrath> k, will do.
[14:30:18]   <mmcgrath> sounds good xDamonx
[14:30:26]   <xDamox> ok c ya all later
[14:30:32]   <mmcgrath> The rest of the stuff on this list is pretty much where it was last week.
[14:30:35]   <xDamox> Ill email by tommorow about pootypedia
[14:30:42]   <xDamox> c ya
[14:30:47]   <mmcgrath> monitoring: nagios is ready but we're waiting for lockbox to get ready
[14:30:58]   <mmcgrath> IDWeb: Sopwith is working on that.
[14:31:00]   ***  xDamox has left chat #fedora-admin ("Leaving").
[14:31:20]   <mmcgrath> Metrics, anyone know anything else about metrics and bouncer?
[14:31:46]   <ja8sun> I have not had much of a chance to work on it
[14:31:57]   <mmcgrath> seems like its been a busy time for everyone.
[14:32:10]   <ja8sun> emailed mike morgan, but no response
[14:32:29]   <mmcgrath> how much longer should we wait on bouncer before we start looking at other options?
[14:32:42]   <ja8sun> I also need to email elliot on whether he has talked t mike about the changes
[14:32:46]   <ja8sun> to bouncer
[14:33:20]   ***  Sopwith has joined chat #fedora-admin.
[14:33:28]   <ja8sun> other then that, not much else to tell..
[14:33:33]   <Sopwith> Sorry guys, just realized I'm very late...
[14:33:34]   <mmcgrath> k sounds good
[14:33:38]   <mmcgrath> sopwith: yooo sopwith!
[14:33:44]   <ja8sun> sopwith: hey
[14:33:52]   <iWolf+ hola.
[14:34:04]   <mmcgrath> we just got to metrics and bouncer
[14:34:30]   <mmcgrath> sopwith: how's IDWeb going?
[14:34:34]   <Sopwith> I even had an alarm set, and just dismissed it and went back to hacking for "just a few more minutes" :)
[14:34:54]   <f13> my alarm stopped working.
[14:35:03]   <Sopwith> idweb is going OK. Bill Baronius thinks it is so close to being finished, but I'm not so sure. :)
[14:35:23]   <mmcgrath> heh.
[14:35:24]   <Sopwith> idweb is not as high a priority now, for reasons that will hopefully become clear soon.
[14:35:39]   <mmcgrath> mysterious!
[14:35:52]   <mmcgrath> ja8sun anything you need to talk to Sopwith about for bouncer?
[14:36:18]   <ja8sun> yeah, have you talked with Mike Morgan on the changes to bouncer?
[14:36:32]   <Sopwith> Not recently.
[14:36:40]   <Sopwith> Did he get back to you about the account?
[14:37:01]   <ja8sun^ no, but I was not sure if I had blind-sided him either
[14:37:19]   <ja8sun> it has been a busy week and I had not followed up with him
[14:38:04]   <ja8sun> sopwith: Was he cool with us making changes to the code?
[14:38:09]   <Sopwith^ OK, I'll try pinging him now.
[14:38:17]   <Sopwith> He seemed very enthusiastic about having anyone interested at all.
[14:38:28]   <ja8sun> sopwith:cool :)
[14:39:02]   <mmcgrath> xDamonx is going to mail to the list with whats going on with pootypedia
[14:39:12]   <Sopwith> I just /msg'd morgamic about it and his reply was "I'll look at it right now"
[14:39:12]   <mmcgrath> ssl is still waiting for fp.org
[14:39:20]   <mmcgrath> awesome.
[14:39:32]   <Sopwith> Stacy is working on fixing RHN access for the PHX machines
[14:39:44]   <mmcgrath> is PXE all set up now?
[14:39:46]   <ja8sun> sopwith: wasn't sure on what kind of information he needed from me
[14:40:14]   <ja8sun> to grant access
[14:41:21]   <Sopwith> And you can now mount the netapp to get the install tree directly (instead of going through HTTP on bastion)
[14:41:41]   <Sopwith> mmcgrath: No, I'd like RHN set up for that (although I suppose we could install the packages manually)
[14:42:00]   <mmcgrath> Sopwith: sounds like fun
[14:42:47]   <mmcgrath> moving the accounts and cvs is still stalled is it? (item "rhlinux.redhat.com migration")
[14:43:42]   <Sopwith> Yup
[14:45:23]   <mmcgrath> whats the word on the translator's compendium?
[14:45:36]   <Sopwith> ja8sun: Maybe to get started, you should do an anonymous checkout and that way you'll have the code to hack on as much as you like.
[14:45:44]   <Sopwith> morgamic: Just trying to get roadblocks outta your way :)
[14:46:10]   <ja8sun^ I have checked out the code, but have not had the time to work with it
[14:46:22]   <Sopwith> OK
[14:46:26]   <ja8sun^ should have some time this weekend
[14:46:30]   <Sopwith> Cool :)
[14:46:44]   <Sopwith> It's spring break here next week, so I hope to have a bit more spare time to stick into this stuff
[14:47:36]   <mmcgrath> There's really not much else on the list: account system rewrite.
[14:48:01]   <mmcgrath> I don't think anyone has been doing much work on that though?
[14:48:31]   <Sopwith> nope, although someone (was it warren?) mentioned to me the need to allow accounts to list multiple e-mail addresses
[14:48:53]   <mmcgrath> just curious, what for?
[14:50:14]   <Sopwith> Mainly because people have been using different e-mail addresses for their bugzilla accounts versus their Fedora accounts
[14:50:20]   <Sopwith> And we need to be able to tie the two together.
[14:50:28]   <mmcgrath> got'cha. I'm actually guilty of that.
[14:51:35]   <mmcgrath> Welp thats the schedule's list. Anyone have anything else?
[14:52:11]   <Sopwith> Just got a report of cvs-via-ssh not working - trying to fix now
[14:52:46]   <mmcgrath> sopwith: I planted the seed for the next review day and OTRS. We'll see if anyone actually reviews it.
[14:52:53]   <Sopwith> xinetd[30413] : FAIL: ssh service_limit from=10.8.34.200
[14:52:53]   <Sopwith> Cool
[14:53:14]   <Sopwith> xinetd[30413] : FAIL: ssh service_limit from=10.8.34.200
[14:53:14]   <Sopwith> doh
[14:54:00]   <mmcgrath> more firewall issues?
[14:54:23]   <mmcgrath> I know proxy4 can't resolve dns right now.
[14:54:28]   <mmcgrath> its in the logs too:
[14:54:34]   <mmcgrath> Mar 2 13:47:11 proxy4 sendmail[1808] : k1Q912tF028544: to=admin@fedoraproject.org, ctladdr=<root@proxy4.fedora.phx.redhat.com> (0/0), delay=4+11:46:09, xdelay=00:00:00, mailer=esmtp, pri=9750867, relay=mx3.fedoraproject.org., dsn=4.0.0, stat=Deferred: Connection timed out with mx3.fedoraproject.org.
[14:55:19]   <mmcgrath> Well aside from a few broken systems, I'll close the meeting unless anyone has any objections or anything to add?
[14:55:22]   <mmcgrath> 30 seconds.
[14:55:38]   <mmcgrath> 15
[14:55:44]   <mmcgrath> 10
[14:55:50]   <mmcgrath> 5
[14:55:51]   <mmcgrath> 4
[14:55:52]   <mmcgrath> 3
[14:55:53]   <mmcgrath> 2
[14:55:53]   <mmcgrath> 1
[14:55:57]   <mmcgrath> Meeting end.